This proposal is designed to burn almost all MEV income for proposers. The auction at the observation deadline is in a way a slot auction masquerading as a block auction. Builders have incentives to bid up to, and even above the expected value of the MEV for the entire slot, since they stand to receive a kickback if they win. A winning builder will update their payload after the observation deadline and provide a small tip so that the proposer includes the updated payload in the slot. Builders that do not win will presumably have worse opportunities to extract MEV and on top of that cannot receive the kickback. They therefore do not have incentives to raise the payload base fee after the observation deadline, because if they then win the proposer auction, they will take a loss.
Now, there will of course be many cases where builders hold back a little in the auction (perhaps to compensate for the prospect of not having a clear winner) or where some late burst of MEV comes in, etc. But it is still reasonable to expect a tempering of block proposals just after the auction, with rather small bumps to the payload base fee, if any, and then potentially some final bidding closer to the end of the proposal auction. Therefore, it seems to me that this effect, while existing, should overall not be that significant in turns of value (at least in the version with an auction design). Especially since the auction can be positioned slightly before the point where the base fee floor is set if necessary. If you can raise the base fee just after the auction, you could just have raised it before, or at least will not be able to raise it by that much without the prospective of taking on a loss.
MEV burn does not materially alter the situation, merely our perception of it. It is not possible today for vanilla builders to build good enough blocks to receive the available MEV value. This is what has led to vanilla building being rather uncommon. Implementing MEV burn does not remove the ability to build blocks for anyone, and does not substantially alter the real economic consequences of such a decision. As a comparison, if the subsidy is bumped for stakers to keep the yield the same before and after implementing MEV burn, then, over time, the outcome for vanilla builders will essentially be the same under the current situation and with MEV burn. The potential “donation” from the vanilla builder to transacting users is not altered in substance.
Of course, our perception of the two situations may be very different, but this is mainly a question of educating users. The current vanilla-builder situation is like an employee who donates the bonus that their employer randomly hands them once in a while. The MEV-burn situation would then be vanilla builders as an employee with a slightly higher salary who donates the difference whenever they once in a while see a poster for a charitable cause. Now if that charitable cause only accepts donations of one million dollars, then the employee may need to avoid donating this one time. We can certainly expect prospective vanilla builders to incorporate a check on the payload base fee floor relative to what they can extract themselves, to ensure that the decision to self-build will not affect them too negatively. Variability may in this way prevent vanilla builders from building their blocks in some cases, if a big MEV opportunity arises. But we must then remember a big advantage of MEV burn—that variability for the more typical solo staker is removed, which is a big win. In many cases, we also specifically would like to burn that big MEV opportunity anyway, to prevent it from falling into the hands of the next proposer.
If the base reward factor is not bumped, such that the yield falls with MEV burn, then the required “donation” will be a larger proportion of rewards than before MEV burn. But that is an effect of an underlying change to the issuance policy, something which is a separate conversation. Reducing the base reward factor right now without MEV burn would have a similar effect on vanilla builders. They’d be forced to forego a larger proportion of their rewards.
I will provide a more extensive write-up on the proposal in a short while.
If builders indeed collude, then vanilla building becomes very cheap. Not sure if I am missing something there.